Swish Building Products is upgrading most of its IT infrastructure. Brian Tinham examines its exemplary choices and expectations
Improvements across the board – inventory reductions, better stock accuracy, more responsive customer service and operational efficiencies (machine utilisation, scrap reduction, maintenance scheduling) – those are among chief expectations at PVC roofline products manufacturer Swish Building Products from its ERP redevelopment project. It’s early days yet, and no-one is prepared to release figures, but Swish management’s thoughts, plans and decisions are well worth some attention.
Just painting a brief picture of the business and its production, the firm turns over £30 million at its Tamworth manufacturing site, which has around 200 employees. It has 21 lines of extruders producing around 50 tonnes of a mix of plastic products a day across the ranges, each equipped with packaging at the end of the lines. The site also has its own tool-making and maintenance shop to service the factory. Altogether it’s a sizeable operation, running 24/7 on three shifts and having to plan and schedule for hundreds of orders a day from its European distributor network.
The firm’s main business IT was – and indeed still is – a now ageing and also incomplete Forth Shift ERP system, with the classic islands of information. Production planning, for example, is stand-alone because, like so many, after the trauma of its Phase One ERP implementation five years ago, Swish hadn’t gone on to install the planning element, instead building a separate planning database. Four years ago, a Preactor finite capacity planning system had been added – but even that was only used as a visual information aid for planners on the shop floor.
The system needed expansion, integration and upgrading to provide the familiar range of operational and inter-departmental business improvements, and a foundation for the company’s plans in term of linking better and more proactively with its customers, distributors and suppliers, and driving a smarter, more flexible organisation.
An obvious course of action might have been to upgrade to the latest release of Fourth Shift. But the project team, headed by IT manager Alan Thompson, reasoned that systems have moved on in terms of capability, ease of use and the rest, and wanted to look at alternatives, particularly from the MRP point of view, as well as customer management, e-business and future-proofing.
“This is my sixth ERP implementation,” says Thompson. “And you need to look at your business objectives and put them first.” So late last year, he kicked off the review – including all the company’s ‘best of breed’ legacy systems outside the ERP itself – with improving everything from planning to customer service on the agenda. Thompson admits to a certain ERP enthusiasm, and so had the advantage of greater exposure and understanding than many. Nevertheless, he did due diligence on several packaged systems from the likes of Exel, InforSwan, IFS, Intenia, K3, Lilly Sofrtware, McGuffie Brunton and Navison, now with Microsoft Business Solutions – as well as the incumbent system’s latest release.
“I started from the position of wanting advanced planning and scheduling to match what for us will be a real planning opportunity,” he says. “But I found that while they’re all fine for some scenarios, ours is unique, and best dealt with by tightly integrating Preactor. Beyond that, they could all have done 85% of what’s required, but I was most impressed with Navision.” And that’s what he recommended, and is now underway.
You see why when he enthuses: “Navision can be whatever we want it to be: it’s highly configurable by using basic built-in functions of the code. You can create business rules almost on the fly where in the past you had to get coders in. In Navision I could see ways to do what we’d been doing before with Microsoft Access and Visual Basic workarounds, but automatically, as part of the core system. It was going to be a much smaller overhead.” And he adds: “We could use ODBC links to the SQL database so we’d get dynamic data to and from other systems – not extracts from batch runs at midnight – so they all become totally trustworthy. There wouldn’t be any islands of information anywhere… Anybody in any department would be able to get to real time information and make better decisions.”
Which would pave the way for his better efficiency and customer service goals. In fact, Thompson’s project scoping, feasbility and plans for the future are exemplary. And he’s just as unequivocal about his other choices: “We’re going for DPS for the vehicle routing and scheduling for our fleet delivery service, with GPS-based vehicle tracking from Minor Planets. Also, for financial analysis we’ve chosen Data Tracker.” And the bottom line: “When that’s all seamlessly integrated, we’ll have the business system we need to go forward for the foreseeable future.”
Back to today and the ERP, and Swish is planning to use the lion’s share of Attain v3.6, including, for example, sales order processing, despatch, finance, production, MRP, warehouse management and marketing via a linked data warehouse. Specifically on the production side, it’s also going for what’s now Preactor advanced planning and scheduling – fully integrated into the main system. Thompson reckons it’ll be live in the third quarter of this year. “We’re currently running the conference room pilot; it’s company-wide, with a representative on the project team from every department.”
Richard Morris, Swish’s operations manager explains that, in production, the plan is to run MRP in Attain ideally twice daily to deal with the mix of products and order volumes, deriving total procurement and production requirements in the usual way – taking into account forecast, orders, stock and work in progress. However, demand will be put into Preactor, which will optimise factory schedules against its finite capacity rules base, providing works orders back via the ERP. Meanwhile, DPS will be run dynamically in real time to handle the distribution side.
Morris sees Preactor’s main role here as ensuring speed and accuracy in production, optimised according to “the factory floor facts of life”. It will, for example, be designed to automatically route products that can only be made on certain lines, recognise limited tool numbers, and look for efficiencies like pairing products for production on one line simultaneously with subsequent splitting operations.
Says Thompson: “Preactor is fantastic value for money; the competition seems to offer not a lot more for much more money… When it’s been integrated we expect to get very big benefits. We expect to be able to reduce materials stock levels by better planning. We’ll also be able to manufacture much smarter: we’re creating pragmatic rules to give us rough cut plans – getting the know-how from the planners and supervisors and putting that into the model.
“We’re not really capacity constrained; our issue is to maximise around how should we be creating orders – the mix of production. We’re looking to automate processes where we have to date relied upon individuals. We’ll still want them to do the work: they’ll manipulate the schedules for best product mixes, react to changes and events and so on. But it will take out the drudge and offer better on time in full delivery – so customer service.”
He also expects the system to take some of the heat off forecasting: “We’ll have pragmatic rules in place to react to orders. It’s much better to do real world stuff instead of relying too much on history… If we make more or less what we made last year we could save big money.”
Then on the customer service side, he says: “I’d say one of my key objectives was that sales will be able to take orders and advise stock availability and production dates if there are shortages at the point of order entry. We aim to get ATP from the links into ERP inventory management, and capable to promise (CTP) from the integration with Preactor production scheduling, so that the system can advise sales executives right on the sales order entry screens with actuals, not just manufacturing forecast information.”
Meanwhile, Morris adds that linking maintenance management with Preactor and Navision will also help, by managing maintenance scheduling against production plans, and bringing critical spares purchasing and inventory management into the main system to further reduce operating costs – and probably improve uptime.
As a general point though, it’s worth noting that this kind of breadth only happens if you’re prepared to put effort and thought in up front. “You do have to look carefully at mapping the business,” says Thompson. “The difference between what people think happens and what actually happens is often surprisingly great.” And there’s change management, people and processes to consider. “You need to find what have become the non value-add processes and look for opportunities to add new value – for example, providing more customer feedback, with information at the point of order, and at despatch. The keys are seamless integration and business analysis in light of what that enables you to achieve.”
And the result: a picture of significant improvement and opportunity all round. Morris says it all: “I’m expecting improvements immediately because of the operational changes we’ll make as a result of the visibility of information. Sales will be able to see production; production shift managers will be able to see what’s supposed to go on the line next, and the following make-to lists… Production booking will be better – on and off as the jobs are done, rather than at the end of the shift. And recording that information from the shop floor will get the business aware of the real world straight away.”
Thompson agrees that shop floor systems development will play an important part in enabling this. “We’ll be doing that as part of the core system at Phase Two, hopefully going live in Q2 2004, with barcoding and automated audit trails after production, warehousing and distribution have been bedded down.”
Benefits there, he says, will include no more documents washing through the system as labels on products get swiped and the ERP system automatically transfers stock from work in progress (WIP), for example, to warehouse stock location, automatically creating shipping transactions and the rest. “We’ll be looking to track materials through production, as well as doing lot traceability and doing inventory creation straight away so that we get real time WIP valuation.”